A trajectory of curiosity
I've been interested in boats that use the Wing In Ground Effect for a
couple decades. The low-energy soaring of pelicans near waves, the
ground effect on a frisbee floating instead of landing after gliding
down near to the ground, and my ground-school teacher explaining
ground effect to cousin Rosanna and me in 2001, have inspired me to
think about this subject. The Russians call it the "screen", or
"ekran", as though there is a kind of screen along the surface of the
ground or water, that lets you through and partly doesn't let you
through, as you're landing.
Even before that, my friend Norbert Wu's father, who was an
aerodynamics professor at Georgia Tech, had explained to me once about
vortex generation under bird wings (as they flap horizontally back
during take-off a vortex starts to roll backward, and then they can
flap forward on top of that vortex to get a lot of extra lift), and
how commercial aviation is trying to figure out how to use that
tremendous lifting power. Cool!
One day, driving home from ground school with Rosanna, I started thinking of
designs of a boat with wings.
My first thoughts had water propellers, and
a canard wing in front (which some have said
must be a BS idea because I don't have a degree in aeronautical
engineering and therefore couldn't know, and don't have the right to
say, anything about the subject; we'll let that pass). My latest
thoughts have obsessed over the Coanda effect
and blown wing tops with
At every step it seems like the secret design is but a few steps
away. Maybe that will be the true eternal take-away from this
fascinating area, but meanwhile, it seems like we are just a few steps
away from cracking the puzzle, and I love that. Please join me as I
put some pieces on the table, and invite you to consider fitting them
together in different ways.
Models to Follow
Instead of describing a lot of bad ideas first, let's start with some
good ones, and present them by example. Here:
RCTestFlight: a WIG model
I love this guy's
effect vehicle over snow
. I can't find his actual name to give credit but his
YouTube ID is RCTESTFLIGHT (!oh! it's Daniel something). You might say
his experiments are primitive but I say they are brilliant and moving
things in the right direction.
First, it looks to me like a Coanda effect
design. Second, in blowing onto the wing upper surface, his
design rejects PAR
, correctly in my opinion.
Third, he has a wingtip flying in water
with little detriment.
Edderitz: a fast boat
I'm also quite impressed
, I'll call it the 2015 Edderitz boat.
For high on-the-water speed, I like its wide-set front sponsons and
rear drive. A balanced WIG boat should work well as a high-speed boat
also, with stability and efficiency operating in full water
Consider as a goal for design a WIG boat with the abilities of the
2015 Edderitz boat, which keeps stably close to the surface. The
Edderitz boat does a great job balancing the two forces of air pushing
it down to the surface and water keeping it above the surface. A
similar balance, even more difficult to achieve, would be between the
non-lift at a higher elevation due to being out of ground effect, and
the lift at a lower elevation due to ground effect. For each velocity
there is some optimum, and if stability can be achieved, more than
stability but a set of parameters where given the velocity then the
elevation is specified and there are strong forces pushing both up and
down to keep the vehicle at that elevation. That's the goal. The
Edderitz boat at 1/8 to 1/2 wingspan elevation. How can we achieve
that? Fly by wire? "Pre-tensioned" hydraulic actuators following a
balance point? What are the forces, can we quantify them and design to
float at the right points? That's what I'd like to see, and I offer
the following puzzle pieces to help us all get there.
Let's start with the bad ideas, then the questionable ideas, then what
seem to me to be the good ideas.
Death by center-of-lift variation.
Ok here are some ways to hopefully not die.
If you are just designing a fast boat, then, like Edderitz,
- make sponsons in front (or
face-plant crash with sponsons not in front) and
- make them aerodynamically neutral (or
backflip crash with sponsons in front but not aerodynamically
In our case (WIG boats), the basic thing is that if your ground
effect is from air stagnation under the wing, then the lift comes
primarily at the trailing edge of the wing where all that air is
finally maximally crushed under the trailing edge, and that pushes
the vehicle up. The air at the front part isn't compressed yet, so
it doesn't push upward that much. So stagnation effect lift is
located emphatically rearward. Then as soon as the vehicle rises out
of stagnation ground effect, the air isn't nearly as compressed at
the trailing edge, and the whole wing contributes more equally to
the total effect, irrespective of front or back, so the center of
lift moves forward. Therefore once the vehicle catches some lift to
get 1/4 or more of a wingspan up, or especially if in the back it
gets up even just a little, then the relative compression at the
trailing edge drops off dramatically. Compare the thin edge of air
under a 5-percent elevation-to-span ratio scenario, for example,
what will happen to the air compression ratio at the trailing edge
if the wing rises by half a span? Instead of getting all the air
crushed into 5% of the space at the very rear edge, it barely
compresses at all under an edge with a 55% gap; the relative
compression is much less between main undersurface wing area and
trailing edge area. Thus a bit away from the ground, the extreme
tail edge concentration of lift shifts toward the center of the wing
chord. Then when the center of lift suddenly distributes forward to
the center of the wing, because you were balanced on the rear-edge
center of lift, now suddenly the whole thing flips up and backwards
Nice way to die, is what I'm saying. This is a problem.
PAR: Power Assisted Ram
No! Don't Blow It Up! I love it
that RCTestFlight's design places the prop
above the wing, because higher speed air is lower pressure so more
lift up there. These designers with PAR (power assisted reinjection
of air under the wing) are creating the wrong effect!
- Thus I don't much like the jets blowing
air under the wings for initial liftoff. Alexeev spent a
lot of time on this one. I don't think he got it right. Why?
- Doesn't fast air above slow air below represent the basic
idea of lift on a curved-top airfoil? Bernoulli, anyone?
Then why blow fast air underneath the wing? It can
only work in stagnation, when the rear edge of the wing is
approximately underwater, then you're blowing up a very
leaky balloon with a jet engine, sure it'll blow up a bit,
but it'll spray water everywhere like crazy and it'll barely
work. (See the
Amphistar, from 8:42 with prop wash directed sharply
similarly.) Mostly blowing between wing and water
amounts to sucking the wing down to be close to the water.
Delifting, not lifting. Lifting only at the barest
beginning, getting the wings just out of the water. Maybe
that's an argument for super low wings, though, as you see
Alexeev's planes were pretty flat bottomed. But I am
reminded of the old two-balloons experiment: blow between
them (fast air) and the balloons bang together, bang bang
bang bang, a little of the pressing apart happens, but
mostly a lot of pulling them together happens. And then
it's a repetitive banging thing. (See the balloon effect,
and see a couple of seconds of the test video for Alexeyev's
Sea Monster design that demonstrates Bernoulli
at 3:37-3:40) Which we don't want, neither the pulling
two surfaces together with fast air rushing between them,
nor the repetitive banging. So for me for now, let's scratch
the PAR (power assisted ram) idea, okay? You wiser people,
please persuade me otherwise. Meanwhile, let's forget it.
- PAR is also unnecessary for the purpose of getting the wings
out of the water if the design starts with wings already out
of the water. Then normal propulsion can be used, and any
hydrodynamic design will allow smooth acceleration up to the
point of liftoff. For example, have a high wing on a shoulder
blade, have a water prop pushing at low speeds and jointly
with air propulsion during liftoff acceleration, then let it
spin out and crank up out of the way after liftoff.
- You might even want to force the beast to stay in water
contact until past the point of minimum liftoff speed, in
order to make use of the water propulsion up to a higher
speed, then let it pop up out of the water, letting the
water propulsion system shut down when there is enough power
in the air propulsion system to keep it flying. What would
that take, eh?
External take off power
- Here's a moonshot inspired, otherwise bad, idea. The much-higher
power requirement for take-off might be remedied by
externally-applied acceleration: a slingshot launch method.
Imagine a weak WIG-boat that can putt-putt along the water, or
fly above it, but that can't make the transition without help.
It would be safe but not convenient, because if you fall out of
the sky you can still putt-putt along to wherever you need to
go. Eventually. No, let's skip this idea, in service of
self-rescue capability, and thus of general functionality. I'll
call it a bad idea.
Naming is a struggle for this category of most fascinating vehicles.
It has the worst names: ekranoplan or screen-plane in Russian, WIG,
AGEC, WISES, these names suck.
I don't have a solution, though I don't mind airboat or wingboat.
Anyhow W-In-G sounds more like WING than WIG to me. Whatever, just
use all the names in your web pages so everyone can find it, whatever
you want to call it. I hate it that we're stuck with WIG Boat.
Not Obviously Bad Ideas
Canard guides underdraft
Early in thinking about WIG designs, it seemed to me that a small wing
out front, a canard, could generate a downdraft below the main wing,
producing greater pressure there and greater lift.
Fig. 10, photo of the Odessa Institute of Merchant Marine Engineers'
WIG boat under Yu. A. Budnitsky.) If the problem of a WIG boat is
packing some more airflow under the wing at a given speed, then a
canard could help.
I have gotten some ridicule for this idea, but later I noticed that
competition racing hydroplane boats use
design. Between the front sponsons and the body is commonly a short
wing with a gap behind it whereby air flowing over this first wing can
get sucked under the following, longer-chord (body) wing or boat
undersurface, and thereby give more of a cushion to ride on.
Also canards are seen in the "Japanese WISES" design
here and the
Wing design here and
So it's not clearly a bad idea.
Water or air propeller?
Perhaps my first WIG idea was the intuition that a water prop might be
more efficient than an air prop because it's pushing against an
incompressible and dense material -- water -- which ought to be better
for mechanical energy transfer than pushing against a compressible
light material, air. Years later I found that N. I. Belavin
14 (cached copy here) reports that
about double the thrust comes from water props as compared with
A stupid idea here is this: WIG boats have a limited safe velocity
range, so it might be an advantage if the power mechanism disengages
when the boat flies up too high, and the prop comes out of the
water. I don't think so: we want a design that provides actual control!
Finally, the efficiency of a high-speed boat prop depends on the blade
being half in and half out of the water, a very very precise elevation
requirement, which even medium waves will prevent. So my thinking has
shifted to air propellers, which can be efficient too, and even their
inefficiency in pulling in side channel air is a benefit in
the Coanda effect much discussed here.
- Consider a vehicle that handles equally with air propulsion,
lift, and control surfaces and with water propulsion, lift, and
control surfaces. Maybe two engines, two propellers, one for
each medium. Maybe air wings AND hydrofoils, both. Maybe an
air rudder like the giant Jorg tail
AND a water rudder and front-corner sponsons. Effects in each
medium should be coordinated, but can help each other during
transitions, for example, the vehicle will be lifted by
flotation then hydroplaning then aeroplaning with a weighted
mixture of effects adding to the total results for lift, and
similarly for propulsion and control.
- Compromise designs can lose on both fronts, but I would
like to see this for myself rather than give up on the
basis of mere principle. Evidently specific distributions
of wave height and separation should be the basis of
design, and should be practically enforced as hard limits
on actual operation, particularly as to take-off.
- I mentioned water ruddering above. Air rudders have to be
enormous to be effective: see Jorg. Maybe we do need some
combination with a water tail rudder for directional control,
perhaps with a fly-by-wire adjustment algorithm or other control
system making the turn smooth irrespective of the
moment-to-moment variation in the amount of rudder water
contact. An Edderitz boat with poor directional control would
be scary. But a tiny amount of water ruddering would resolve
that issue completely. If it can be done safely, stably in
potentially wave-filled environments, and without much drag.
- Considering much slower speeds, a likely bonus feature needed
for widespread, consumer type use might be a water prop for
in-the-water putt-putt movement, parking and getting out into
the main ways, as in marinas and close-in boating areas where
the giant winds of an air propellor might not be neighborly.
Double the thrust for the same horsepower, sure why not. You
might need a separate motor, or else some efficient means of
transferring work from a front air prop to a rear water
propeller (like a separate electric motor). Yes, if the boat
can hike itself into the air it ought to have the propulsion to
move around a marina, but we might also need to consider the
neighbors part of the time.
- Thus some form of bimodality may be needed.
Wing/body shape controls
- I can imagine a few mechanisms for modifying wing/body parameters:
- One is a "shoulder blade", a mechanism for raising the
wing from low-wing (in WIG flight) to high-wing (in
flotation). I don't like the wing stuck in the water
during taxiing or docking, it should be high. Some
rotating truss design should be able to lower the wing
once airborn (to reduce wave battering on the body at a
given wing elevation above the surface).
- Another is a parking mechanism, rotating a single-span
wing back, or folding two spans on each side against
each other like a bird. The bird model is attractive
if probably unrealistic, but should be understood:
after landing, birds will fuss around and fold the
wings under, up, and back until finally at a low-energy
rest position. Upon takeoff, they sweep wings backward
generating a horizontal-axis vortex with the top
rolling toward the tail of the bird, followed by a
forward sweep making use of the added lift produced by
flying over a vortex that is rolling under your wing.
This is Dr. Wu's point.
- Many such adjustments such as folding multi-span wings
could require quite manipulable wing shapes, and in
turn the feather concept has its value, as feathers
overlappingly form a conjoint surface of manipulable
thickness and shape, each very light and controlled at
the attachment end by shrugs and stretches in
coordination with the whole group to form the wingform
of current utility, depending on landing, lifting off,
soaring, etc. Of course bats fly happily without
feathers, depending on a skin stretch factor that is an
alternative for us also in this design space.
Water contact can be more than a drag. If wave height is normally
distributed, then occasional freak waves are inevitable, and the
design of WIG boats should allow for non-catastrophic water contact
during normal flight. This means knifing sponsons in the front
corners where the bounce off the water is pretty soft.
At the same time, sponsons must be aerodynamically neutral
Curran's Lehigh University thesis on aerodynamics of high-speed
sponsons has influenced my thinking: the front left and right
water-contacting corners of a WIG boat should be a bit like Curran's
Sponson A, that is, aerodynamically neutral in varying angles of
attack, but hydrodynamically efficient, with a step to reduce wetted
area at speed.
- I take it as proven that high speed
water-contacting vehicle designs including WIG boats must have
front right and left corner sponsons which are aerodynamically
neutral while providing hydrodynamic lift.
- I take it as desireable that sponsons provide
mixed-medium stabilization with smooth rather than abrupt
impact from free flight at some range of angles of
entry. Therefore considered from bottom upwards, they should
begin quite small and thin in displacement volume and in wetted
surface area, perhaps knife-like vertical or
designed to impact the water during flight creating only
minimal drag and medium lift, at low wave penetration heights.
Their water displacement volume should remain small up to
fairly deep penetrations as through wave tops or to stabilize
flight when rolling onto one side or the other, and get larger
in displacement only when the sponson is operating as an actual
Flying with a wing in water contact
I tend to think a WIG boat should be comfortable in mixed media. That
is, not just putt-putting in water and flying in air, but as a fast
boat able to fly, and as a flying vehicle in more or less
simultaneous contact with the water. Why? Because practical WIG use
will certainly have to deal with the transition, in which both water
contact and flight simultaneously occur, and if that transition is a
happy stable state, then good, whereas if not, then we have danger and
RCTestFlight's vehicle shows success
with at least one wing corner in water contact. I'd like to see
both as much as neither. It likes to fly in a large curve with a
single corner in water contact. For increased stability, I say,
turn the corner sponson/pontoon into a knife
edge so it has minimum drag, tweak it so it tracks straight, if
possible, with either one or both front corners in water contact
during normal cruising flight speeds.
- Comparing the two sources of ground effect, span-dominated
versus chord-dominated, the main effect is span-dominated. Van
Opstal's graph shows that at a height of 5% of wingspan,
the drag is reduced by to a 30% fraction compared with free
flight. Whereas at a height of 5% of the chord, the
drag reduces only from 1.1 to 0.8 as compared with free flight.
5% of the CHORD! It would require a chord longer than the
average wingspan to get far less than half the effect, if my
reasoning is correct. Therefore ground effect is hugely
dominated by the wing span effect. Another way to say this is
that wing downwash is a bigger energy sink than wingtip
vortices. And therefore making a long-chord wing, which many
designs are based on including Lippisch and Jorg, is a mistake;
whereas a short-chord, longer span wing, is the way to maximize
Witness low-soaring birds such as seagulls and pelicans: Long
If the Alexeyev ekranoplan designs have short wings, it's
because the lift is so great they don't need more.
From Lippisch to Coanda
Consider a sequence of proposals taking off one from the other:
- Lippisch by his designs said, spread
out the rear edge in a front-back dimension by making the rear
edge sharply diagonal forward as the wing reaches farther out, so that a good part of the
"rear" edge is quite a bit forward of the rearmost part of the
rear edge. He also added a small pontoon wing also up front,
to add more front-loaded lift to the mix. Therefore the change will
be less when the nose kicks up.
- But following Van Opstal's findings,
consider de-emphasizing the stagnation (chord-dominated) ground
effect, and instead emphasizing span-dominated ground effect lift
(where ground effect efficiency comes from reduced wing
downwash in proximity to the ground surface) or plain lift or a
blown wing top in your design.
- A compromise shape might use BOTH a blown wing top with prop
above and long chord along the centerline of the vehicle body
-- AND at the same time a long span. Think Lippisch X-112
without the delta shape, just a long chord for a middle section
of wing, while emphasizing extra long pontoon wings.
- In idle imagination I wonder about a biplane, with Coanda lift
on an upper, long-chord, short-span wing, perhaps containing
the passenger/cargo spaces within the wing itself, all above a
lower, long span wing using ground effect lift. Was this
connected to the thinking behind original (1920's) biplane
designs with a shorter lower wing and longer upper wing? When
the prop blows harder, the plane will lift hard and could even
rotate forward like an accelerating helicopter. With reduced
prop power, ground effect on the lower wing still keeps you
- In excited imagination I conceive a wing of variable,
adjustable, controlled aspect ratio.
- In pancake configuration, it lifts strongly,
artificially, at low velocity with a long chord,
- In pelican configuration, it lifts strongly,
naturally, with a short chord, long span, efficiently
propellored wing with little or no Coanda blown
- In mixed configuration, wing tips ("fingers"), extend
before wing inner sections ("elbows"); doing so they
grab more lift as velocity and still-air bernoulli
effect permit; as they begin to bite, more of the wing
tips can extend, eventually sacrificing Coanda-captured
surface area and the lift therefrom,as strong and
reliable pelican lift safely permits.
- Propellor downwash orientation can be maintained so
that lift can be quickly reasserted by wing retraction
into Coanda (downwash, captureable by wing entering
within the) space.
- Possibly, if safe, prop orientation may be
straightened, up, closer to path-parallel, when even
over the central body, Coanda lift is no longer needed.
- Cellularly connected, adjustable x/y/z ratio, jointly
but not identically controlled, wing structural units,
may be jointly designed to achieve arbitrary end
configurations and transitional shape patterns under
one or few control variables, similar to
morphing. Cells may have constant volume or contents,
etc, but be flexible (in at least certain designed
dimensions) surfaces; I hate to suggest rubber since
that limits temperatures of operation. Cell shape
adjustments may be via motorized, hydraulic, or
pneumatic power delivery and by central or distributed
controls. Cells may also carry wing surface forming
features (e.g., feather surfaces) to influence shared
wing shaping and to maximize lift in the then current
Coanda effect lift
All this is consistent with Coanda lift as being responsible for the
wonderful out-of-water lift and stable flight properties of
RCTestFlight's vehicle, and a key puzzle piece in the search for the
WIG boat of the future.
- I believe the excellent lift and easy-liftoff qualities
of RCTestFlight's vehicle are due
to the Coanda effect whereby the outflow of the propeller above
the wing entrains onto the upper surface of the wing, hence,
due to the Bernoulli effect, provides extreme lift to the wing
in addition to forward propulsion.
- In his videos, 1 and
you can see the propeller axis points somewhat downward toward
the top of the wing surface, thus making it easier for the
Coanda entrainment to occur.
- RCTestFlight's vehicle's wing has an extremely long chord. Why?
In Coanda/Bernoulli lift enhancement the propeller outwash
entrains onto the wing upper surface. The total lift force is
proportional to the amount of wing surface to which the
resulting lowered pressure applies. So you could increase lift
by increasing the upper surface area impacted by the outwash,
that is essentially by either increasing width (span) or length
(chord) of the wing.
For more, read
flaps, especially look for the phrase Upper Surface Blow,
which in the one built case, the YC-14, increased the
coefficient of lift to 7, (compare with a Boeing 747 at high
altitude cruise having a coefficient of lift of 0.52 according
- Increase the wing span, and nothing happens, because the
outwash is centered on the propeller and can only spread
laterally a limited amount. You could increase the amount
of the span which is affected by using multiple
propellers, or by providing ducts or vanes to spread the
propulsion wash laterally across the wing. Ugh. Or
- Increase the wing chord (increasing the duration of
airflow contact with the wing), and you directly increase
the area effected by Coanda/Bernoulli. Thus more lift.
The effect occurs over a larger area with a longer
above-wing wash entrainment surface. It lowers the
pressure above the wing over a larger total surface area.
- So the outwash is essentially a narrow (though widening)
but long resource for imposing lift onto a wing. Actually
its shape is likely somewhat triangular, with greater span
influenced as the wash reaches along a greater distance of
Coanda 2: Blown wingtop observations and discussion
- So far, a blown wingtop design seems quite promising. Make the
prop wash entrain to the top of the wing via the Coanda effect.
Then it should will produce huge lift, according to the Bernoulli
effect. 14x maybe. Shouldn't that be of some assistance
especially during take off? This seems to have been part of
Lippisch's thinking in the X-112, which has a propellor located
so most of its outwash is above the wing. Yet the X-113 and
X-114 which should be advances have the motor rearward and
higher, so little wingtop entrainment and Coanda/Bernoulli lift
can occur. Noticeably the X-112 seems to spend a lot more time
at >1/5 wingspan elevations where the later models seem to fly
lower, tighter to the water -- that is, reliant more on
chord-based ground effect lift. The X-112 flies away from the
- I just had an idea. Suppose you could vastly increase
the chord of the wing to provide a larger blown surface and more
lift during take-off, but at the same time vastly decrease the
chord of the wing to spend more of the propellor energy on
propulsion instead of lift. Overlapping "feathers" (slidable
overlapping wing surfaces) on a moving understructure, to shorten
and lengthen the chord. The biplane mentioned herein is an
alternative. Lippisch's pontoon wings added to a stagnation
ground effect reverse delta wing is also a way to deliver both
effects in one vehicle.
- The blown wing top concept is supported by
Channelwing, which increases wing-top air velocity by
creating a venturi effect within a half-channel over the wing,
and demonstrates vertical take off and 8-13 lbs lift per
- The blown wing top concept is also consistent with what my
impressions of watching the library of Youtube videos on WIG
boats. Some designs seem to fly at 1/5 to 1/2 wingspan and can
lift out of ground effect, though laboriously; I will call these
the "floaters". Other designs fly with the training wing edges
barely clearing the water, say 1/20 to 1/10 wingspan, and these
cannot lift out of ground effect; I will call these the
"huggers". The floaters include Lippisch X-112, Japanese WISES,
and RCTestFlight. The huggers include Lippisch X-113/114 and
derived forms, and Jorg. Huggers might be able to float if they
get substantial extension wings outside their pontoon/sponsons,
which give much more lift. Floaters also have blown wing tops
(with the propellor outwash all or primarily on top of the wing),
while Huggers seem to blow under the wings. Am I making my
Thank you for your interest and patience with this inventory of
partially digested and integrated concepts. I invite you to share
with me your thoughts, questions, corrections and friendly
suggestions. And consider, if you are interested in this area, how
might we help each other advance the state of this art?